The routing layer for modern AI models.

Metapi acts as a centralized proxy and management platform for your LLM API endpoints. Control rate limits, monitor usage, and route requests safely.

Secure Proxy Layer Conceal underlying provider keys and endpoints behind a unified, token-based proxy interface.
Usage Telemetry Monitor token consumption, latency, and request volume across all your active models in real-time.
Dynamic Routing Seamlessly switch upstream providers or load-balance requests without modifying client application code.

Sign in

Access the Metapi dashboard

$ curl -X POST https://meta.mer3y.xyz/v1/chat/completions \
  -H "Authorization: Bearer sk-..." \
  -d '{"model": "claude-3-opus", "messages": [{"role": "user", "content": "Hello"}]}'

HTTP/2 200 OK
{"id": "msg_01...", "model": "claude-3-opus", "choices": [...]}

$ HTTP/2 200 OK
{"id": "msg_01...", "model": "claude-3-opus", "choices": [...]}

Built for developers, designed for scale.

Everything you need to manage, secure, and optimize your AI API infrastructure in one place.

🔑

Key Management

Centralize all your provider API keys. Rotate, revoke, and audit access without touching a single line of client code.

📊

Real-time Analytics

Track token usage, request latency, error rates, and cost breakdown across every model and provider in real-time dashboards.

⚡

Rate Limiting

Set granular per-key, per-model rate limits. Protect upstream providers from abuse while ensuring fair usage across your team.

🔄

Load Balancing

Distribute requests across multiple provider endpoints. Automatic failover ensures zero downtime when a provider goes down.

🛡️

Access Control

Fine-grained permissions per API key — restrict which models, endpoints, and features each consumer can access.

🔌

OpenAI Compatible

Drop-in compatible with the OpenAI API format. Switch your base URL and start routing — no SDK changes required.

50+

Supported Models

99.9%

Uptime SLA

<50ms

Proxy Overhead

∞

Scalable Endpoints

One endpoint, every model.

Point your existing OpenAI SDK to Metapi. We handle the routing, failover, and telemetry.

# Just change the base_url — that's it.

from openai import OpenAI

client = OpenAI(
    base_url="https://meta.mer3y.xyz/v1",
    api_key="sk-your-metapi-key",
)

response = client.chat.completions.create(
    model="claude-sonnet-4-20250514",
    messages=[
        {"role": "user", "content": "Hello!"}
    ],
    max_tokens=1024,
)

print(response.choices[0].message.content)

Start routing in minutes.

Get your API key, point your SDK, and let Metapi handle the rest. No credit card required.

Get Your API Key →